NeuroResource

Spatiotemporal Control of

Opioid Signaling andBehavior

Highlights

d A light-sensitive mu-opioid-like receptor (opto-MOR) was

generated and characterized

d Opto-MOR initiates canonical MOPR signaling both in vitro

and in neurons

d Photoactivation of opto-MOR in selected GABAergic neurons

induces reward or aversion

d Opto-MOR is a novel tool for in vivo optodynamic control of

opioid signaling

Siuda et al., 2015, Neuron 86, 923–935May 20, 2015 ª2015 Elsevier Inc.http://dx.doi.org/10.1016/j.neuron.2015.03.066

Authors

Edward R. Siuda, Bryan A. Copits, ...,

Robert W. Gereau IV,

Michael R. Bruchas

Correspondencebruchasm@wustl.edu

In Brief

Siuda et al. develop a photosensitive mu-

opioid-like receptor (opto-MOR) that

triggers cAMP inhibition and MAP kinase

activation, couples to GIRK currents, and

internalizes like the mu-opioid receptor.

Photostimulation of opto-MOR within

discrete GABAergic nuclei induces real-

time preference or aversion.

Neuron

NeuroResource

Spatiotemporal Controlof Opioid Signaling and BehaviorEdward R. Siuda,1,2,7 Bryan A. Copits,1,3,7 Martin J. Schmidt,1,7 Madison A. Baird,1,6,7 ReamAl-Hasani,1William J. Planer,1

Samuel C. Funderburk,1 Jordan G. McCall,1,2 Robert W. Gereau IV,1,2,3,4 and Michael R. Bruchas1,2,3,4,5,*1Department of Anesthesiology, Basic Research Division, Washington University in St. Louis, St. Louis, MO 63110, USA2Division of Biological and Biomedical Sciences, Washington University School of Medicine, St. Louis, MO 63110, USA3Washington University Pain Center, Washington University in St. Louis, St. Louis, MO 63110, USA4Department of Anatomy and Neurobiology, Washington University in St. Louis, St. Louis, MO 63110, USA5Department of Biomedical Engineering, Washington University in St. Louis, St. Louis, MO 63110, USA6Present address: University of Washington, Department of Pharmacology, Seattle, WA 98195-7280, USA7Co-first author

*Correspondence: bruchasm@wustl.edu

http://dx.doi.org/10.1016/j.neuron.2015.03.066

SUMMARY

Optogenetics is now a widely accepted tool forspatiotemporal manipulation of neuronal activity.However, a majority of optogenetic approachesuse binary on/off control schemes. Here, we extendthe optogenetic toolset by developing a neuromodu-latory approach using a rationale-based design togenerate a Gi-coupled, optically sensitive, mu-opioid-like receptor, which we term opto-MOR. Wedemonstrate that opto-MOR engages canonicalmu-opioid signaling through inhibition of adenylylcyclase, activation of MAPK and G protein-gatedinward rectifying potassium (GIRK) channels andinternalizes with kinetics similar to that of the mu-opioid receptor. To assess in vivo utility, we ex-pressed a Cre-dependent viral opto-MOR in RMTg/VTA GABAergic neurons, which led to a real-timeplace preference. In contrast, expression of opto-MOR in GABAergic neurons of the ventral pallidumhedonic cold spot led to real-time place aversion.This tool has generalizable application for spatiotem-poral control of opioid signaling and, furthermore,can be used broadly for mimicking endogenousneuronal inhibition pathways.

INTRODUCTION

Opioid receptor-targeting drugs have been used as analgesics

and recreationally abused for hundreds of years. Due to a lack

of effective alternatives, they remain on the front lines for acute

pain management, severe anti-tussive treatment, and other indi-

cations despite their high abuse potential. Mu (MOPR), kappa

(KOPR), delta (DOPR), and nociceptin opioid peptide receptors

(NOPR) persist at the forefront of basic science and drug discov-

ery efforts on disorders ranging from gastrointestinal ailments to

pain, addiction, and depression (Al-Hasani and Bruchas, 2011).

Among these receptor systems, the mu-opioid receptor has

been the most intensely studied due to its long-established

involvement in analgesia, euphoria, and reward in response to

morphine-like analogs. However, despite years of study, opioid

research had been limited by a fundamental challenge. Studying

the specific effects of opioids with spatial, temporal, and cell-

type-specific control is virtually untenable, especially in the

context of the CNS. Systemic drugs bind MOPRs on heteroge-

neous cell populations across multiple brain regions. MOPR

expression at both pre- and post-synaptic sites within overlap-

ping discrete regions precludes determination of circuit level

contributions. Therefore, approaches to selectively limit engage-

ment of MOPR signaling to restricted cell populations with tem-

poral control that closely mimics endogenous opioid kinetics is a

first key step toward unraveling these issues and developing

future therapeutic strategies where opioids are the most effec-

tive treatment regimen.

Recent developments in optogenetics and molecular biology

provide an ideal strategy for addressing these questions. Class

A G protein-coupled receptors (GPCRs), including the rat

rhodopsin receptor and MOPR, have structural and functional

similarities that can be exploited to create hybrid receptors

with unique functional properties. More specifically, the light-

sensitive external portion of rhodopsin receptors can be com-

bined with internal signal transduction components of other

GPCRs to produce so-called opto-XR chimeras capable of initi-

ating and terminating receptor-specific signaling events with

temporal precision enabled by pulses of light (Airan et al.,

2009; Gunaydin et al., 2014; Kim et al., 2005; Masseck et al.,

2014). Furthermore, packaging these receptors into Cre recom-

binase-dependent viruses using loxP flanked doubled inverted

open (DIO or FLEX) reading frames allows for restricted expres-

sion in discrete cell types within isolated brain regions yielding

spatial control of GPCR signaling in vivo (Atasoy et al., 2008;

Zhang et al., 2010).

Here, we present the generation and characterization of a new

photosensitive mu-opioid-like chimeric receptor (we term opto-

MOR). We show that opto-MOR suppresses cyclic AMP (cAMP)

levels, activates MAPK signaling, and internalizes in a similar

time course to nativeMOPR. Furthermore, it functionally couples

Neuron 86, 923–935, May 20, 2015 ª2015 Elsevier Inc. 923

to GIRK channels in GABAergic neurons of the periaqueductal

gray (PAG) and rostromedial tegmental region (RMTg),

mimicking properties of native MOPR (Blanchet and Luscher,

2002; Ingram et al., 2008; Matsui et al., 2014; Vaughan et al.,

2003). Finally, we demonstrate here that opto-MOR promotes

classical MOPR-mediated reward and aversion behaviors in

distinct brain circuits. Together, these findings establish opto-

MOR as a spatiotemporally precise MOPR analog and support

its utility for studying opioid circuitry and behavior.

RESULTS

Developing a Chimeric Photosensitive Receptor withComponents of the Mu-Opioid ReceptorIn order to effectively generate an optically sensitive MOPR-

based receptor chimera that would have high probability of

structural similarity to mu-opioid receptors and retain native

opioid Gi/o-protein signaling, we used the closest class

A-related receptor as a backbone. Rat rhodopsin RO4 has

been previously shown to couple to Gi/o signaling pathways

in vitro and thus served as the ideal template for generation of

an opto-MOR (Hille, 1994). Using the constraint-based multiple

alignment tool (COBALT, NCBI; http://www.st-va.ncbi.nlm.nih.

gov/tools/cobalt/re_cobalt.cgi), rat rhodopsin RO4 was aligned

against rat MOPR to identify and align transmembrane, intracel-

lular, and extracellular domains (Figures 1A and S1A). To confer

MOPR-like coupling, we isolated the rhodopsin RO4 sequence,

retained the critical photoisomerizing RO4 retinal binding site

(Figures S1A and S1B), and inserted the rat MOPR intracellular

loops and C terminus (Figures S1A and S1C). A protein folding

prediction was modeled by bioinformatics software (Roy et al.,

2010) in order to accurately project how our various chimeras

would result in a seven transmembrane protein with matched

features (Figures 1A andS1D). This approach allowed for optimal

conservation of Gi/o receptor dynamics, a high degree of photo-

sensitivity, while simultaneously providing the critical intracel-

lular communication components for engaging mu-opioid

signaling pathways in vitro and in vivo.

Opto-MOR Is Photosensitive and Effectively MimicsMu-Opioid Signaling DynamicsIn order to determine if opto-MOR is functional and traffics to the

membrane, we first examined whether a YFP-tagged opto-MOR

expresses at the cell membrane in a pattern similar to that of the

wild-type MOPR-GFP-tagged receptor. We found that both re-

ceptors express at the plasma membrane in a similar manner

(Figures 1B and 1C) in unstimulated cells. Opioid receptor char-

acterization has demonstrated that agonist stimulation of opioid

receptors causes interactions with inhibitory G proteins (Gai) to

induce inhibition of cAMP production (Childers and Snyder,

1978; Al-Hasani and Bruchas, 2011) (Figure 1A). In order to

determine whether cells expressing opto-MOR couple to the

same canonical MOPR signaling pathways, we assessed their

response to blue LED (465 nm, 1 mW, 20 s) stimulation in a

battery of well-known in vitro opioid signaling assays in direct

parallel with wild-type rat MOPR using the high affinity, selec-

tive MOPR agonist, [D-Ala2, NMe-Phe4, Gly-ol5]-enkephalin

(DAMGO; 1 mM). First, we show that opto-MOR maintains mem-

924 Neuron 86, 923–935, May 20, 2015 ª2015 Elsevier Inc.

brane expression levels similar to that of MOPR in unstimulated

HEK293 cells (Figures 1B and 1C). Second, we found that photo-

stimulation of opto-MOR and DAMGO activation of MOPR

caused a time-locked decrease in forskolin-induced intracellular

cAMP levels with similar time constants of activation (Figures

1D–1F, S2A, S2B, S2D, and S2E). To verify the specificity of

our constructs, we show that opto-MOR does not respond to

the selective MOPR full agonist DAMGO, and likewise wild-

type MOPR does not respond to photostimulation (Figures

S2C and S2F). Additionally, we show that opto-MOR is maxi-

mally activatedwith 465 nm light and shows less efficacy at other

wavelengths in cAMP inhibition (525, 630, and 660 nm) (Fig-

ure 1G). Finally, we found that opto-MOR is highly sensitive to

light and requires very little light power for photoactivation (Fig-

ures 1H and S2G), while varying LED pulse lengths resulted in

similar levels of cAMP inhibition (Figure 1I). Opto-MOR and

MOPR caused similar levels of cyclic AMP inhibition via photosti-

mulation and agonist treatment, respectively, suggesting that

opto-MOR couples to canonical mu-opioid signaling pathways,

yet utilizes rapid time-locked photoswitching to engage Gai-

mediated inhibition of adenylyl cyclase signaling.

Agonist stimulation of all four opioid receptors has been shown

to recruit various factors resulting in mitogen-activated protein

kinase (MAPK) activation (Bruchas et al., 2011; Al-Hasani and

Bruchas, 2011). MOPR has been shown to elicit a rapid initial

peak in the phosphorylation of extracellular signaling-regulated

kinase (pERK) in neurons, astrocytes, and transfected cell cul-

tures (Belcheva et al., 2005). Here, we examined whether opto-

MOR and MOPR produce similar kinetics and efficacy in

engaging pERK signaling in HEK293 cells. In complementary ex-

periments, we found a rapid and transient increase in pERK

(�2–5 min) in response to blue LED photostimulation of opto-

MORandDAMGOapplication toMOPR-expressingcells (Figures

1J–1L). pERK returned to basal levels 60–90 min after either pho-

tostimulation or DAMGO treatment. Furthermore, opto-MOR-

mediated activation of ERK was mostly independent of LED

pulse time (Figure S2H) and only mildly affected by light power

(Figure S2I), suggesting that time-locked photoactivation of

opto-MOR immediately engages the MAPK signaling cascade.

Opioid receptors are well known to be rapidly regulated by

arrestin-clathrin-mediated internalization pathways. To assess

whether opto-MOR exhibits similar activation-induced receptor

regulation and engages canonical mu-opioid receptor internali-

zation machinery, we performed side-by-side experiments

whereby we compared the kinetics and efficacy of LED or

agonist-stimulated receptor internalization on opto-MOR or

MOPR. Following photostimulation or DAMGO treatment,

respectively, both opto-MOR and MOPR internalized rapidly

(within 15 min) with similar rates of internalization (tin), with a

peak effect at 30 min after agonist or light stimulation (Figures

2A, 2B, S3A, and S3B). Opto-MOR receptors remained punctate

in the cytosol at all post-stimulation times tested, lasting up to an

hour. Opto-MOR internalization over this time course is consis-

tent with previously reported wild-type MOPR internalization by

both widely used synthetic and endogenous ligands (Al-Hasani

and Bruchas, 2011).

Agonists such as etorphine and DAMGO and endogenous

opiate peptides such as Met-enkephalin and b-endorphin have

0 400 800time (sec)

0

20

40

60

80

100

cAM

P(%

max

) - LED+ LED

0 400 800time (sec)

- DAMGO+ DAMGO

opto-MOR MOPR

0 30 60 900

20

40

60

80

100

time (min)

pER

K(%

max

)

0 30 60 90time (min)

opto-MOR MOPR

D E

0 2 5 15 30 60 90pERK

actin

pERK

actin

opto-MOR

MOPR

4442

42

min4442

42

0 5 15 30 60 90 min

J

RhoMOPR

A

cAMP

Gi/o arrestin

pERKpERK

opto-MOR

optoMOR

MOPR

F

0

10000

20000

30000

40000

50000cA

MP

(AU

C)

** **

G

465 525 630 6600

50

100

150

LED (nm)

cAM

P(%

cont

rol)

****

**

H

K L

optoMOR

MOPR

B Copto-MOR

MOPR

2 5 15 600

50

100

150

light stim (sec)

cAM

P(%

cont

rol)

I opto-MOR

-6 -5 -4 -3 -20

50

100

150

power (log mW)

cAM

P(%

cont

rol)

opto-MOR

opto-MOR

0

20

40

60

80

%at

mem

bran

e

Figure 1. Opto-MOR and MOR Share Ca-

nonical Intracellular G Protein Signaling

Mechanisms

(A) Schematic of opto-MOR showing activation of

cAMP and pERK pathways.

(B) Opto-MOR and MOPR are highly expressed at

the membrane in HEK293 cells (scale bar, 10 mm).

(C) Opto-MOR (purple, n = 43 cells) and MOPR

(n = 37 cells) share similar surface expression in

HEK293 cells.

(D) Optical stimulation (blue bar; 20 s, 1 mW)

of opto-MOR reduces forskolin-induced (dashed

line) cAMP in HEK cells (n = 9–10 experiments).

(E) Application of DAMGO (gray bar; 1 mM) reduces

forskolin-induced (dashed line) cAMP in HEK293

cells expressing MOPR (n = 3 experiments).

(F) Light and DAMGO significantly reduce for-

skolin-induced cAMP in opto-MOR (purple; n = 12)

and MOPR (gray; n = 3).

(G) cAMP inhibition by opto-MOR is most efficient

at 465 nm (n = 6–8 experiments; *p < 0.05, **p <

0.01, ***p < 0.001 via one-way ANOVA with Bon-

ferroni’s multiple comparison test).

(H) cAMP inhibition by opto-MOR is power

dependent (n = 3–9 experiments).

(I) cAMP inhibition by opto-MOR is not dependent

on light pulse length (n = 2–10 experiments).

(J) Representative immunoblots show similar ki-

netic increases in pERK in response to light (1 min,

1mW) or DAMGO (1 mM) in opto-MOR (purple) and

MOPR (gray)-expressing cells.

(K and L) Quantification of light-induced pERK in

opto-MOR (n = 6) (K) and DAMGO-induced pERK

in MOPR (n = 2) (L). Data are represented as

mean ± SEM. See also Figures S1 and S2.

been shown to cause MOPR internalization 30 min after drug

exposure, as receptor endocytosis and recycling to the mem-

brane reaches steady state around 30 and 60 min, respectively.

To test whether opto-MOR displays functional desensitization in

Neuron 86, 923–

addition to internalization, we photosti-

mulated opto-MOR and induced inhibi-

tion of cAMP (Figures 2C and 2F). In

parallel, we photostimulated opto-MOR

expressing HEK293 cells at saturating

power (1 mW, 20 s), waited varying de-

grees of time, and then photostimulated

again (i.e., pre-pulsed). We found that at

approximately 15 min following pre-pulse

stimulation, there is a significant loss of

opto-MOR function, with no difference

observed from the unstimulated control

cells (Figures 2D and 2F). However,

longer time periods between the two light

pulses show that within approximately

60 min there is restoration of opto-MOR

function in inhibition of cAMP (Figures

2E and 2F). However, it is unclear whether

this restoration is due to receptor recy-

cling or de novo receptor synthesis. To

address this, we repeated this experi-

ment but pretreated cells with either brefeldin A, which blocks

translocation of proteins from the ER to the Golgi and has

been used in studies of MOPR recycling, or Dyngo-4a, an inhib-

itor of dynamin that prohibits receptor internalization and has

935, May 20, 2015 ª2015 Elsevier Inc. 925

C D

E F

B

A Figure 2. Opto-MOR and MOPR Internalization and Recovery from

Desensitization

(A) Representative images show internalization of opto-MOR (colorized yellow;

scale bar, 50 mm) and MOPR (inset, gray; scale bar, 10 mm) expressed in

HEK293 cells in response to light and DAMGO. 0, 5, 15, 30, and 45 min time

points are represented. Arrowheads show examples of internalized receptor.

(B) Quantification of receptor internalization in opto-MOR (purple; tin = 8.9 min;

n = 16–43 cells per time point over 2 experimental replicates) andMOPR (gray;

tin = 8.5 min; n = 24-38 cells per time point over 3 experimental replicates) in

response to light and DAMGO-induced activation, respectively.

(C) Opto-MOR inhibits forskolin-induced (dashed line) cAMP following light

stimulation (n = 4 traces).

(D) A second light pulse 15 min following the first shows a loss of opto-MOR

activity (n = 3–4 traces).

(E) cAMP inhibitory activity returns to baseline levels 60 min following an initial

light pulse (n = 3 traces).

(F) Time course of recovery from desensitization (n = 3–11 replicates; ***p <

0.001 via one way ANOVA followed by Dunnett’s multiple comparison test to

control). Data are represented as mean ± SEM. See also Figure S2.

926 Neuron 86, 923–935, May 20, 2015 ª2015 Elsevier Inc.

recently been shown to prevent intracellular GPCR cAMP

signaling. Following a 1 hr interpulse interval (Figure S3D), we

found that pretreatment with brefeldin A did not prevent opto-

MOR-induced inhibition of cAMP (Figures S3C and S3E).

Furthermore, Dyngo-4a pretreated cells also show cAMP inhibi-

tion, suggesting that opto-MOR continues to inhibit cAMP from

the plasma membrane with no loss of signal when inhibiting dy-

namin-clathrin pathways (Figures S3C and S3F). These data

together suggest that opto-MOR may indeed recycle and use

clathrin-dynamin endocytosis, which has been documented in

opioid receptors (Gupta et al., 2014) and other GPCR systems

(Irannejad et al., 2013; Tsvetanova and von Zastrow, 2014),

and that opto-MOR may have some utility in dissecting these

diverse signaling dynamics.

Opto-MOR Expresses and Is a Functional Receptor inDRG NeuronsTo test whether opto-MOR is expressed and functions outside

the context of an artificial heterologous system, we generated

a Cre-dependent adeno-associated viral construct (Figure 3A),

whereby we inserted opto-MOR into a double inverted open

reading frame flanked by LoxP and Lox2227 sites (Zhang

et al., 2010). Utilizing this virus (AAV5-EF1a-DIO-opto-MOR-

YFP), we infected cultured dorsal root ganglia (DRG) neurons

from Advillin-Cre+ mice (da Silva et al., 2011; Zhou et al.,

2010). We did not detect any signal in uninfected control neurons

(Figure 3B). In Advillin-IRES-Cre+opto-MOR DRG neurons, we

observed opto-MOR expression in as little as 2 days post-trans-

duction (green, Figure 3C), suggesting rapid Cre-mediated exci-

sion in DRG. Strong opto-MOR expression was seen after 5 days

(Figure 3D), which partially co-localized with synapsin (Fig-

ure 3D), demonstrating that opto-MOR is trafficked efficiently

along neuronal processes and to presynaptic terminals. We

also show that following photostimulation of opto-MOR-ex-

pressing DRGs, there is an increase in pERK compared to un-

treated controls (Figures 3E and 3H). Following photostimulation

(1 min, 1 mW), we observed an increase in pERK levels 5 min

following an initial light stimulation that returns to basal levels

within 30 min (Figures 3F–3H). We also report evidence for pho-

tostimulated receptor internalization in these DRG cultures

A Cre-dependent double inverted open (DIO)Adeno-Associated Virus (AAV)

lox2722

Avil promoter Cre5’ 3’

EF1α

YFP opto-MOR 5’3’ WPRE

loxP

AAV5

AAV5 EF1α opto-MOR YFP WPRE 5’ 3’

2 DIV

DAPIopto-MORsynapsin

DAPIopto-MORsynapsin

5 DIV

DAPIopto-MORsynapsin

B

C

D 5 DIV

2 DIV

5 DIVuninfected

5 DIVuninfected

XATG

0 5 30 600

20

40

60

80

100

time (min)

pER

K(%

eYFP

)

***

opto-MOR

DAPIopto-MORpERK

pERK pERK

pERK

DAPIopto-MORpERK

DAPIopto-MOR

pERK

E F

G H

0 min 5 min

5 min

Figure 3. Opto-MOR Expression and Function in DRG Neurons

(A) Opto-MOR packaged into an AAV5-DIO-EF1a-YFP viral construct. Cre

expression is driven from the sensory neuron-specific Advillin promoter

following insertion after the initial start codon, resulting in Cre-mediated

recombination and inversion of the opto-MOR construct into the appropriate

50-30 orientation.(B) Uninfected DRGs colabeled with the nuclear stain DAPI (blue) and

synapsin (red).

(C) Opto-MOR expression (green) in DRGs 2 days in vitro (DIV). Large arrow-

heads point to the soma of infected neurons.

(D) Opto-MOR expression (green) in DRGs 5 DIV. Small arrowheads highlight

opto-MOR expression in synaptic terminals colabeled with synapsin (red).

Scale bars for (B)–(D), 30 mm.

(E) Unstimulated opto-MOR expressed in DRGs after 5 DIV labeled with pERK

(red and inset).

(F) Opto-MOR expressing DRGs (5 DIV) 5 min following photostimulation and

labeled for pERK (red and inset). Open arrows denote internalized punctate

receptors.

(G) Higher magnification of opto-MOR expressing DRGs (5 DIV) 5min following

photostimulation and labeled for pERK (red and inset). Open arrows denote

internalized punctate receptors.

(H) pERK intensity normalized to opto-MOR-YFP intensity at varying times

following photostimulation (n = 3–32 DRG neurons per time point; ***p < 0.001

via one way ANOVA followed by Dunnett’s multiple comparison test to 0 min.

Scale bars for panels (E)–(G), 20 mm. Data are represented as mean ± SEM.

(Figures 3F–3H), suggesting that the receptor can be dynami-

cally regulated in physiologically relevant systems. These data

closely mimic the kinetics of ERK activation seen in HEK293 cells

expressing opto-MOR (Figures 1J and 1K) and further support

the utility of using opto-MOR to dissect signaling in neuronal

preparations.

Opto-MOR Couples to Gbg-Mediated GIRK Currents inPAG GABAergic NeuronsIn order to assess whether opto-MOR couples to similar intracel-

lular signaling pathways following endogenous MOPR activation

in neurons, we focused on the periaqueductal gray (PAG),

which contains a high proportion of MOPR+ neurons (Vaughan

et al., 2003). We injected AAV5-EF1a-DIO-opto-MOR-YFP into

vGAT-IRES-Cre+ mice (Vong et al., 2011) within the PAG

(AP: �5.0 mm, ML: ± 0.5 mm, DV: �2.8 mm) and targeted

YFP+ GABAergic neurons for whole-cell recordings from acute

brain slices 2–3 weeks after injection. Photostimulation with

470 nm LED light (10 mW/mm2) evoked rapid outward currents

together with a simultaneous decrease in input resistance that

was reversed by bath application of barium (1 mM), a blocker

of GIRK currents (Figures 4A and 4B), suggesting that opto-

MOR activation leads to downstream activation of GIRK chan-

nels, which is consistent with the activation of endogenous

MOPR receptors in these neurons (Vaughan et al., 2003). A par-

allel set of experiments was performed onMOPR positive cells in

the PAG. Bath application of DAMGO (1 mM) showed similar out-

ward currents and decreased input resistance that were also

barium sensitive (Figures 4C and 4D). Voltage ramps performed

pre- and post-stimulation demonstrated current-voltage curves

consistent with the properties of GIRK channels, which were in-

hibited by barium (Figures 4G and 4H). Peak changes in holding

current and input resistance for opto-MOR (LED) and MOPR

(DAMGO) were not statistically different following stimulation,

or in the presence of barium (Figures 4E and 4F), suggesting

Neuron 86, 923–935, May 20, 2015 ª2015 Elsevier Inc. 927

DAMGO barium

0 5 10 15 20 250

1

2

3

4barium washout(norm to pre-light)

time (min)

+ 470 nm LED + bariumopto-MOR positive

MOPR positive + 1 μM DAMGO + barium

-120 -100 -80 -60 -40 -20-50-25

0255075

100

opto-MOR GIRK current

opto-MORLEDbarium

0.00.20.40.60.81.01.21.4

-2 0 2 4 6 8 10time (min)

0

5

10

15

20

25

0500

10001500200025003000

0.00.51.01.52.02.53.0

A B C D

E F G H

I

J

K

L

pre-light

pre-light

0.0

0.5

1.0

1.5

2.0

2.5bariumbarium

bariumbarium

LED

LED

LED

LED

0

10

20

30

40DAMGO barium

0

500

1000

1500

2000

2500 DAMGO barium

200 ms20 mV

0.0

0.5

1.0

1.5

2.0

2.5DAMGO barium

0.40.60.81.01.21.41.61.8

-120 -100 -80 -60 -40 -20-50

0

50

100

150

MOPR GIRK current

baselineDAMGObarium

0

5

10

15

20

25 -1500

-1000

-500

0

-20

-15

-10

-5

0 0

500

1000

1500

2000

Post-stimulation

LED DAMGO

1 mM barium

-2 0 2 4 6 8 10time (min)

-2 0 2 4 6 8 10time (min)

-2 0 2 4 6 8 10time (min)

-2 0 2 4 6 8 10time (min)

-2 0 2 4 6 8 10time (min)

-2 0 2 4 6 8 10time (min)

-2 0 2 4 6 8 10time (min)

hold

ing

cure

nt (p

A)

hold

ing

cure

nt (n

orm

)

inpu

t res

ista

nce

(MΩ

)

inpu

t res

ista

nce

(nor

m)

hold

ing

cure

nt (p

A)

inpu

t res

ista

nce

(MΩ

)

hold

ing

cure

nt (n

orm

)in

put r

esis

tanc

e (n

orm

)

LED DAMGO LED DAMGO LED DAMGO who

le-c

ell c

urre

nt (p

A)

membrane potential (mV) membrane potential (mV)

who

le-c

ell c

urre

nt (p

A)

Δ ho

ldin

g cu

rren

t (pA

)

Δ in

put r

esis

tanc

e (MΩ

)

Δ ho

ldin

g cu

rren

t (pA

)

Δ in

put r

esis

tanc

e (MΩ

)

hold

ing

curr

ent (

norm

)

0 5 10 15 20 25time (min)

inpu

t res

ista

nce

(nor

m)

Figure 4. Photostimulation ofOpto-MOR andActivation of EndogenousMOPRsHave Similar Effects onNeuronal Physiology and Excitability

in GABAergic PAG Neurons

(A) Representative plots from an opto-MOR+ neuron in acute PAG slices illuminated with 470 nm LED light (10 mW/mm2) for 60 s. The top trace shows a rapid

outward current in response to illumination, while the bottom trace depicts the simultaneous drop in input resistance. Bath application of 1 mM barium (red

shading) both blocked the outward current and reversed the change in input resistance.

(B) Normalized summary plots showing the response to LED stimulation from additional opto-MOR+ neurons as described in (A) (n = 4).

(C) Example traces recorded from a MOPR+ neuron stimulated with 1 mM DAMGO (gray bars) showing similar outward currents (top) and decreased input

resistance (bottom) compared to opto-MOR+ neurons shown in (A) and (B).

(D) Normalized summary plots from additional MOPR+ neurons depicting the response to DAMGO as described in (C) (n = 4).

(E) Quantification of the peak changes in holding current and input resistance following LED (purple) or DAMGO (gray) stimulation (n = 4).

(F) Quantification of the peak changes in holding current and input resistance after application of 1 mM barium (n = 4).

(G and H) Representative current-voltage traces from a 250 ms voltage ramp from �20 to �120 mV, in an opto-MOR+ neuron (G) before (purple trace) and after

60 s LED stimulation (blue), or a MOPR+ neuron (H) before (black) or after stimulation with 1 mMDAMGO (gray). Both currents were reduced by the GIRK channel

blocker barium (1 mM, red).

(I and J) Left traces depict voltage traces from current-clamp recordings of an opto-MOR+ (purple traces) (I) or MOPR+ neuron (black) (J) in response to hyper-

polarizing current injections of �20 and �10 pA, and depolarizing current injections of 1- and 2-times rheobase. Middle traces show decreased input resistance

and excitability following LED (blue) and DAMGO (gray) in response to the same current injections before stimulation. The right traces demonstrate the increased

input resistance and neuronal excitability observed following GIRK channel block with 1 mM barium (red). Dashed lines indicate �60 mV membrane potential.

(K and L) Normalized summary plots of the persistent outward current (K) and decreased input resistance (L) in opto-MOR+ neurons following light stimulation and

subsequent barium washout (n = 3). These effects were observed >45 min after 60 s LED stimulation. The dashed lines indicate the average response before

illumination. Data are represented as mean ± SEM. See also Figure S4.

928 Neuron 86, 923–935, May 20, 2015 ª2015 Elsevier Inc.

LEDin DAMGO

LEDin DAMGO

LEDin DAMGO

-120 -100 -80 -60 -40 -20-100

0

100

200

300

400

membrane potential (mV)

-3 -2 -1 0 1 2 3 4 50.0

0.5

1.0

1.5

2.0

time (min)

-3 -2 -1 0 1 2 3 4 50.0

0.5

1.0

1.5

2.0

time (min)

inpu

t res

ista

nce

(nor

m)

hold

ing

curr

ent (

norm

)

LEDin DAMGO

-3 -2 -1 0 1 2 3 4 50

5

10

15

20

25

30

time (min)

-3 -2 -1 0 1 2 3 4 50

500

1000

1500

2000

time (min)

hold

ing

curr

ent (

pA)

inpu

t res

ista

nce

(MΩ

)

0

10

20

30

DAMGO LED

hold

ing

curr

ent (

pA)

0

500

1000

1500

DAMGO LEDinpu

t res

ista

nce

(MΩ

)

who

le-c

ell c

urre

nt (p

A)

DAMGOLED

B

D

-20 mV

-120 mV

-60 mV

1 μM DAMGOLED

A

-3 0 3 6 9 12 150

5

10

15

20

25

30

time (min)

-3 0 3 6 9 12 150

500

1000

1500

2000

time (min)

1 μM DAMGOLED

hold

ing

curr

ent (

pA)

inpu

t res

ista

nce

(MΩ

)C

0

10

20

30

40

50

hold

ing

curr

ent (

pA)

DAMGO LED

E

0

500

1000

1500

inpu

t res

ista

nce

(MΩ

)

DAMGO LED

F5 sec LED 60 sec LED GIRK current ramp

Figure 5. Activation of Endogenous Mu-Opioid Receptors Occludes GIRK Channel Activation by Opto-MORs

(A) Example plots from an opto-MOR and MOPR+ GABAergic neuron in the PAG. An outward current and simultaneous decrease in input resistance were

observed following prolonged application of 1 mMDAMGO, which plateaued after�5 min. Brief (5 s) LED illumination did not result in further changes to either of

these parameters.

(B) Example plots from a different neuron positive for both opto-MOR and MOPR in the presence of 1 mM DAMGO. The plateaued response to DAMGO is not

shown in this example. Prolonged LED illumination (blue, 60 s) did not alter the holding current or input resistance.

(C) Normalized summary graph showing the occluded responses to LED illumination following stimulation of endogenous MOPRs with DAMGO (n = 3).

(D) Quantification of the holding current and input resistance in the presence of DAMGO and following 5 s of LED stimulation (n = 3).

(E) Same quantification as in (D), but with 60 s illumination (n = 3).

(F) Representative GIRK channel ramp in the presence of DAMGO and following LED stimulation. Data are represented as mean ± SEM.

that activation of both receptors is coupled to similar down-

stream effectors. We also found that activation of both opto-

MOR and MOPR substantially decreased neuronal excitability

in response to step current injections, which was also reversed

by application of barium (Figures 4I and 4J). We did, however,

notice a persistent outward current and significantly reduced

input resistance following washout of barium (Figures 4K and

4L), indicating that brief activation of opto-MOR can couple to

downstream effectors to suppress neuronal activity for an

extended period of time.

We next asked whether opto-MOR and MOPR activation en-

gages converging intracellular signaling cascades leading to

GIRK channel activation, or whether these two receptor popula-

tions signal through parallel subcellular pathways. To test this,

we recorded from MOPR+/YFP+ GABAergic neurons in the

PAG and applied saturating concentrations of DAMGO (1 mM).

We observed sustained outward currents coupled with a

decrease in input resistance (Figure 5A). After this response pla-

teaued, we stimulated neurons with either a brief LED light pulse

(5 s, Figure 5A) or prolonged illumination (1 min, Figures 5B and

5C).We found that opto-MOR coupling toGIRK channel currents

and changes in input resistance were occluded by sustained

activation of endogenous mu-opioid receptors (Figures 5D–5F),

suggesting that activation of opto-MORs converge onto the

same pool of intracellular signaling cascades as endogenous

mu-opioid receptors. In addition, we injected AAV5-EF1a-DIO-

opto-MOR-YFP into the rostromedial tegmental nucleus

(RMTg) (AP: �3.9 mm, ML: �0.1 mm, DV: �4.5 mm) of a new

cohort of vGAT-IRES-Cre mice and targeted YFP+ interneurons

for whole-cell recordings. Photostimulation with 470 nm LED

light (10 mW/mm2) evoked outward currents together with a

decrease in the input resistance (Figures S4A andS4B), suggest-

ing opto-MOR activation is coupled to GIRK channels, which is

consistent with the endogenous MOPR function in these neu-

rons (Matsui and Williams, 2011; Matsui et al., 2014). Voltage

ramps performed pre- and post-stimulation demonstrated cur-

rent-voltage curves consistent with the properties of GIRK

channels (Luscher et al., 1997) (Figure S4C).

Neuron 86, 923–935, May 20, 2015 ª2015 Elsevier Inc. 929

A

C

G H I J

D E F

B

Figure 6. Photostimulation of Opto-MOR Causes MOR-like Behavioral Profiles In Vivo

(A and B) Mice with AAV5-DIO-opto-MOR-YFP injected into the RMTg (yellow) (A) with a fiber optic implanted into the VTA (red) and the ventral pallidum

(yellow) (B).

(C) Viral expression in VTA and RMTg.

(D) Viral expression in fibers (open arrows) projecting into the VTA (identified by tyrosine hydroxylase staining in red) as well as GABAergic interneurons of the VTA

(closed arrows).

(E) Viral expression in VP.

(F) Viral expression in VP GABA neurons (closed arrows).

(G and H) Mice expressing opto-MOR in the RMTg-VTA (n = 8) display significantly increased real-time preference behavior compared to controls (Cre-

littermates, n = 7) as shown by representative position heatmaps (G) and the mean preference for the stimulation-paired chamber (H).

(I and J) Mice injected into the VP (n = 16) display the converse aversion behavior in the place preference assay compared to controls (Cre� littermate controls,

n = 5) spending less time in the light-stimulated chamber also shown by position heatmaps (I) and mean place preference results (J). *p < 0.05 via unpaired t test.

Data are represented as mean ± SEM. See also Figure S5.

Taken together, these results from two different brain regions

suggest that opto-MOR couples to canonical MOPR G protein

signaling pathways, yet retains rapid photo-engagement of

opioid signal transduction pathways, making it an ideal tool for

spatiotemporal approaches to study MOPR signaling in real

time in vivo.

In Vivo Photostimulation of Opto-MOR in SelectedGABAergic Neurons Mediates Preference or AversionTo demonstrate the utility of opto-MOR for studying opioid neu-

ral circuits in behavior, we locally infused AAV5-EF1a-DIO-opto-

MOR-YFP into the brains of two naive groups of vGAT-IRES-Cre

mice, one group in the RMTg (Figure 6A) and one group in the

ventral pallidum (VP) (Figure 6B), and waited at least 3 weeks

for optimal viral expression at the soma (Figures 6C–6F). Fiber

930 Neuron 86, 923–935, May 20, 2015 ª2015 Elsevier Inc.

optic implants (Sparta et al., 2012) were chronically implanted

above the VTA and the VP, respectively. Mice were placed into

black opaque boxes and allowed to explore both chambers

freely to evaluate preference or aversion behavior using the

real-time place preference assay. During testing, mice received

constant 10 mW 473 nm blue laser light upon entry into the ‘‘light

stimulation’’ side (Figures 6G and 6I). Preference and aversion

were calculated as the difference in time spent in the light-stim-

ulated box compared to the box with no stimulation, with prefer-

ence defined as more time spent in the photostimulated box and

aversion defined as less time in the photostimulation chamber.

vGAT-IRES-Cre+opto-MOR:RMTg mice displayed a significantly

greater preference for the stimulation than their Cre(�)-injected

opto-MOR littermate controls (Figures 6G and 6H, t13 = 2.231,

p < 0.05; Figures S5A and S5B). In addition, based on our data

demonstrating a power response relationship with opto-MOR in

activating cAMP and pERK (Figures 1H and S2I), we posited that

different light powers might produce different behavioral pheno-

types. As such, we focused on real-time place preference in

vGAT-IRES-Cre+opto-MOR:RMTg mice. We found that animals

receiving no light stimulation in the VTA show no preference for

either chamber (Figure S5I). However, when we increase the

power to 1 mW, animals begin to show a place preference and

spend more time in the conditioned chamber than those

receiving no light stimulation, but less time than those receiving

10 mW photostimulation (Figure S5I), thus demonstrating that

opto-MOR signaling can be more finely manipulated in vivo.

In contrast, mice injected with opto-MOR into the VP (vGAT-

IRES-Cre+opto-MOR:VP) displayed significant place aversion to

photostimulation, as compared to Cre(�) littermate controls (Fig-

ures 6I and 6J; t19 = 2.174, p < 0.05; Figures S5E and S5F).

Importantly, there were no differences in locomotor activity

throughout the session for either group (Figures S5C and S5G),

and light stimulation did not induce acute changes in locomotion

in the stimulation chamber (Figures S5D and S5H). Finally, as a

positive control, and to compare the efficacy of our opto-MOR

experiments for a more traditional optogenetic strategy, we

tested a new cohort of mice in which we inhibited GABAergic

neurons in the VTA using halorhodopsin (eNpH3.0) in the real-

time preference assay (Figures S5J–S5L) as previously reported

(Jennings et al., 2013). These data were strikingly similar to

inhibiting RMTg GABAergic neuronal population with opto-

MOR (Figure S5A), suggesting that inhibition of these popula-

tions has comparable behavioral phenotypes with alternative

optogenetic tools. Taken together, these behavioral results indi-

cate that opto-MOR functions to engage signaling and neuronal

inhibition with the ability to elicit robust real-time behavioral

responses.

DISCUSSION

Here we describe the development of a new opto-XR receptor

based on the mu-opioid receptor. We show that this opto-

MOR behaves functionally like its wild-type MOPR counterpart

in engaging canonical opioid G protein signaling pathways and

receptor internalization. Furthermore, we found that selective

expression of opto-MOR in GABAergic neurons within two

distinct brain regions promotes diverse behavioral responses

consistent with those previously observed with local MOPR

agonist infusion. This opto-MOR receptor could potentially

also be used alongside chemogenetic approaches, in peripheral

studies to assess nociceptive circuits, or as a robust way to

inhibit neuronal activity without the limits of the currently used

Cl� and H+ pump approaches that include halorhodopsin or

archaerhodopsin (Berndt et al., 2014; Raimondo et al., 2012;

Wietek et al., 2014).

Opto-MOR Mimics Mu-Opioid Receptor SignalingIn VitroA few other opto-XR receptors have been reported (Airan et al.,

2009; Barish et al., 2013; Kim et al., 2005), including an opto-b2adrenergic receptor as well as the recently described 5HT1A-like

opsin-chimera using the C-terminal components of the wild-type

receptor (Masseck et al., 2014). However, few reports have

directly examined the intracellular signaling dynamics of opto-

XR signaling as they relate to their wild-type receptor counter-

part, nor have these receptors been utilized in native circuits

where their endogenous complementary receptor types are ex-

pressed. This report shows side-by-side that GPCR Class A

rhodopsin-like chimeras can be generated to mimic neuropep-

tide receptor function in vitro. We found that opto-MOR indeed

engages inhibition of cAMP, couples to GIRK outward currents

in neurons, and activates ERK rapidly following light pulse expo-

sure (within 2 min), and finally that opto-MOR internalizes rapidly

in response to photostimulation. Furthermore, we show the

opto-MOR receptors functionally desensitize (Figures 2A and

2B), in a manner similar to that of their wild-type MOPR counter-

parts, suggesting that these receptors couple to native down-

regulation signaling pathways including G protein-coupled

receptor kinases (GRKs) and arrestins. In addition, our data

suggest that similar to other previously reported GPCRs, these

receptors may potentially signal at both membrane and intracel-

lular domains (Irannejad et al., 2013). The possibility therefore ex-

ists that opto-MOR may be capable of optical activation while

localized intracellularly, not only highlighting an important

caveat, but also illustrating the potential utility of this chimera

to elucidate components of intracellular signaling cascades.

However, future studies to examine opto-MOR regulation of

signaling dynamics in native neurons are required, and mutation

of the opto-MOR receptor at C-terminal sites previously associ-

ated with receptor regulation could potentially provide advanta-

geous optogenetic tools in dissecting MOPR-related signaling

and tolerance in vivo.

In Vivo Utility of Opto-MOR for Behavioral OptogeneticStudiesWe primarily generated this unique tool to facilitate mimicry of

opioid signaling in vivo for use in optogenetic studies. The ability

to time-lock activation of mu-opioid-like signaling in an awake-

behaving animal provides an alternate approach for mapping

the role of opioid receptor signaling in discrete neural circuits

and cell populations in real time. In order to demonstrate the util-

ity of this approach in vivo, we chose two brain regions widely

known to express high levels of endogenous MOPR that have

previously been linked to mu-opioid-dependent reward and

aversion. In a first series of experiments, we examined the

effects of opto-MOR expression in VTA/RMTg GABAergic neu-

rons using real-time place testing. Recent work had demon-

strated that mu-opioid receptors are expressed on both local

VTA interneurons as well as a large proportion of nearby RMTg

GABAergic cells. It is hypothesized that these receptors likely

act to disinhibit dopaminergic activity resulting in MOPR-medi-

ated reward behavior. We demonstrate here that photostimula-

tion of opto-MORVGATcre+ receptors in RMTg neurons produces

real-time place preference and reward seeking behavior. This

result is consistent with reports using mu-opioid agonists

into this region (Jalabert et al., 2011; Jhou et al., 2012; Matsui

et al., 2014). In addition, prior reports have demonstrated

that optogenetic activation or inhibition of VTA and RMTg

GABAergic neurons drives real-time place preference or aver-

sion behavior (Tan et al., 2012; van Zessen et al., 2012), further

Neuron 86, 923–935, May 20, 2015 ª2015 Elsevier Inc. 931

suggesting that these VTA neuronal populations are poised to

regulate dopaminergic output and behavior.

The ventral pallidum (VP) region is well known to express mu-

opioid receptors, and regional activation has been reported to

lead to both reward and aversive behavioral responses (Hjelm-

stad et al., 2013). We therefore expressed opto-MOR in VP

GABAergic neurons and selectively targeted coordinates corre-

sponding to the previously reported opioid ‘‘hedonic cold spot’’

(Smith and Berridge, 2005; Smith et al., 2009) in order to deter-

mine if activation of opto-MOR signaling in this region could elicit

real-time aversion. Consistent with prior reports showing that

MOPR agonism in this anterior central region decreases reward

seeking (Smith et al., 2009), we observed that opto-MOR activa-

tion within these neurons produced avoidance of the photosti-

mulation chamber. These two behavioral assays demonstrate

that opto-MOR can elicit robust real-time opioid-like behavioral

responses in vivo and highlight their utility for future studies

examining circuit and cell-type functions of this pathway.

In addition, these in vitro and in vivo findings coupled with our

evidence of coupling to GIRK signaling highlight the utility of

opto-MOR for more generalizable optogenetic inhibition studies,

such as those that are routinely conducted using halorhodopsin

or archaerhodopsin. Opto-MORcouples to native signaling path-

ways to regulate neuronal activity via GIRKs and might be

preferred since it does not come with the potential caveats that

the photosensitive chloride pumps (eNpHR3.0) bring with them,

including alterations in the reversal potential of theGABAA recep-

tor, that lead to changes in synaptically evoked spiking activity in

the period after photoactivation (Raimondo et al., 2012). Future

work utilizing opto-MOR in other neuronal cell types and circuits

is warranted, however, for determining its utility as a broadly

generalizable optogenetic inhibition approach.

Limitations and Future DirectionsAlthough we have developed a photosensitive mu-opioid-like re-

ceptor for the interrogation of opioid neural circuits in vitro and

in vivo, there are some potential limitations of this approach

that are worth noting. We chose to use vertebrate rhodopsin

as a backbone for generation of opto-MOR because it has pre-

viously been shown to couple to Gi/o-related signaling (Li

et al., 2005) and thus made for an ideal chimeric approach. How-

ever, this choice of rhodopsin has limitations since the receptor

is very photosensitive, as indicated in our in vitro studies and

in vivo behavior. This is an important consideration when using

the tool in vitro for slice physiology and circuit analysis, as well

as in vivo, whenwanting to elicit discrete activation in a particular

anatomic subregion. To work around these caveats, we were

careful to use the tool under red light conditions in vitro, and

while in vivo we kept the fiber implants capped. To resolve

this, future variants might be warranted with mutations that

render opto-MOR less photosensitive and/or use other short-

wavelength opsins as part of the chimeric design.

It is also important to note that the receptor is only partially a

mu-opioid receptor and thus might display a different protein-in-

teractome than the wild-type MOPR. Indeed, we show that the

receptor utilizes a pool of intracellular signaling similar to that

of MOPR (Figures 1 and 2) and traffics to synapses (Figure 3);

however, recent work has suggested a complex regulatory pro-

932 Neuron 86, 923–935, May 20, 2015 ª2015 Elsevier Inc.

teome of MOPR interactions (Al-Hasani and Bruchas, 2011), and

thus we are careful to interpret our data in light of known trans-

membrane and extracellular loop residues that are important

for MOPR function in neurons.

Future studies using the opto-MOR in side-by-side experi-

ments with other exciting recent developments (Banghart and

Sabatini, 2012) in optical control of neuropeptide function should

provide researchers with a means to better dissect the role of

opioid neuropeptide signaling in complex neural circuits and in

freely moving mice (Kim et al., 2013). Understanding the relative

dynamics of neuropeptide release and receptor signaling is a

key question in neuroscience, and opto-XRs that mimic endog-

enous signaling pathways hold promise in delineating some of

these mechanisms. Furthermore, there is a growing need to

better understand the molecular-cellular mechanisms of opioid

tolerance in pain, and thus the optodynamic properties of this

approach may have utility in this arena. The use of this tool will

provide an additional layer of spatiotemporal specificity when

used with traditional pharmacology, physiology, and behavioral

approaches.

ConclusionsIn summary, here we provide and characterize a novel tool for

neuroscience that allows for spatiotemporal control of opioid re-

ceptor signaling in vitro and in vivo. Opto-MORs functionally

couple to canonical intracellular signaling pathways and have

utility in dissecting opioid contributions to behavioral effect.

The combination of opto-XR approaches with chemogenetic

and traditional optogenetic channel approaches expands our

toolbox for understanding neuropeptide signaling within neural

circuits, within real-time, freely moving animals. Furthermore,

advances in receptor chimeras such as opto-MOR will facilitate

the generation of additional related approaches for other GPCR

classes.

EXPERIMENTAL PROCEDURES

Construct Design and Cloning

Using the constraint-based multiple alignment tool (COBALT, NCBI; http://

www.st-va.ncbi.nlm.nih.gov/tools/cobalt/re_cobalt.cgi), rat rhodopsin 4

(RO4) sequences were aligned against rat MOR. RO4 extracellular domains,

including the retinal binding site, and transmembrane domains were identified

and conserved. HindIII/EcoRI sites were added to the N terminus while Xhol/

BamHI sites were added to the C terminus of the chimeric sequence. The

stop codons were replaced with leucine residues, and silent mutations were

added to remove an endogenous Xhol restriction digest site 990 bp into the

gene. The construct was synthesized by IDT in a non-expressing pSMART

vector (Lucigen) then cloned without stop codons into pcDNA3-YFP (Invitro-

gen) containing ampicillin resistance cassettes and CMV promoter regions

for expression in mammalian cells. Restriction enzyme digest sites EcoRI

and XhoI were used to excise the opto-receptor construct from the pSMART

vector and purified by gel extraction. pcDNA3-YFP was also digested with

EcoRI, and Xhol and T4 Ligase were used to ligate opto-MOR into the vector.

E. coli was transformed with pcDNA3-opto-MOR-YFP plasmids, plated on

ampicillin-containing plates. GeneWiz verified chimeric sequence integrity.

Rat MOR-GFP was kindly donated from the lab of Dr. Charles Chavkin at Uni-

versity of Washington, Department of Pharmacology (Celver et al., 2004).

Real-Time Measurement of cAMP Dynamics

Stable HEK cell lines containing opto-MOR and MOR were transfected

with the pGloSensor-22F cAMP plasmid (Promega E2301) using JetPrime

(Polyplus) transfection reagent per manufacturer’s instructions. Stable co-

transfected cells were maintained under both G418 (400 mg/ml) and hygrom-

ycin (200 mg/ml) selective pressure. The day before an experiment, cells

were plated on 96-well tissue culture-treated plates (Costar), 60K–100K

cells/well, and allowed to recover overnight at 37�C, 5% CO2. The next day,

media was replaced with 2% GloSensor reagent (Promega) suspended in

CO2-independent growth medium (GIBCO) and incubated for 2 hr at 37�C.For real-time cAMP, a baseline was first obtained with no treatment by

recording relative luminescent units (RLUs) every 6 s for 1 min using a

SynergyMx microplate reader (BioTek). 1 mM forskolin (dashed line in each

figure) was applied to each well, and RLUs were recorded for 3 min. Cells

were then treated with either a 20 s, 1 mW, blue LED pulse or 1 mM DAMGO

(Tocris) and RLUs recorded for an additional 10 min. For data expressed

as cAMP (fold baseline), RLUs for 1 min of baseline were averaged and all

subsequent RLUs were then divided by this average. For data expressed as

cAMP (percent max), raw RLUs were entered into GraphPad Prism (v5.0d,

GraphPad Software), and the normalization function was used to assign the

lowest RLU a value of 0% and the highest RLU a value of 100%. Time

constants were calculated in GraphPad Prism using one-phase association

(Y = Y0 + (Plateau-Y0)*(1-exp(�K*x))) nonlinear regression analyses yielding

a time constant value (ton). For cAMP (percent control) data, area under the

curve for each treatment group was normalized to its corresponding experi-

mental control. Each data point is a minimum of three wells per experimental

replicate.

Immunoblotting

Immunoblotting for MAPK activation was performed as previously described

(Bruchas et al., 2011). Opto-MOR and MOPR expressing cells were grown

to 100% confluency in 24-well plates. Cells were serum starved in plain

DMEM for at least 4 hr before drug or LED treatment using Plexon LED Driver

LD-1 (465 nm). Cells were harvested in 500 ml of lysis buffer containing (in

mM) 50 Tris-HCl, 300 NaCl, 1 EDTA, 1 Na3VO4, 1 NaF, 10% glycerol, 1%

Nonidet P-40, and 1:100 of phosphatase and protease inhibitor mixture set

(Calbiochem, EMD Millipore). Lysates were sonicated for 10 s and spun

down for 20 min (14,000 3 g, 4�C). 25 mg protein was run in each well,

and blots were transferred to nitrocellulose (Whatman), blocked with 5%

FBS albumen for 1 hr, then incubated with 1:1,000 goat anti-rabbit

pERK1/2 (phospho Thr-202/Tyr-204; Cell Signaling) and 1:20,000 goat anti-

mouse b-actin (Abcam) antibodies at 4�C overnight. Membranes were

washed 4 times for 10 min each in Tris-buffered saline and 0.1% Tween-

20 (TBST) and incubated with 1:10,000 IRDyeTM 800- and 700-conjugated

affinity-purified anti-rabbit or anti-mouse IgG in a 1:1 mixture of 5% milk/

TBS and Li-Cor blocking buffer (Li-Cor Biosciences) for 1 hr at room temper-

ature. Membranes were washed 4 times for 10 min each with TBST, then 2

times with TBS to remove Tween. Blots were visualized, background sub-

tracted, and quantified with Odyssey Infrared Imager (Li-Cor Biosciences).

pERK bands were normalized to b-actin from each sample and calculated

as fold increase from baseline or as percent of maximal ERK phosphorylation

as appropriate.

Receptor Internalization and Confocal Microscopy

Opto-MOR-YFP and MOPR-GFP-expressing cells were plated to approxi-

mately 50% confluence on poly-D-lysine-coated coverslips (Becton Dickin-

son). Following LED (1 W/cm2, 1 min) or DAMGO (1 mM) treatment, cells

were washed 3 times with PBS and then fixed with 4% paraformaldehyde.

Coverslips were then mounted with Vectashield Hard Set mounting medium

with DAPI (Vector Laboratories) and imaged on a Leica TCS SPE confocal

microscope (Leica), and Application Suite Advanced Fluorescence (LAS AF)

software calculated receptor internalization (inside fluorescence/total fluores-

cence 3 100).

Animal Subjects

Group-housed adult (25–35 g) male vGAT-IRES-Cre mice (Vong et al.,

2011) and Cre(�) littermate controls were used for all in vivo experiments.

For eNpH3.0 VTA GABA experiments (see Supplemental Materials),

GAD2-Cre mice were used. All mice were maintained on a 12 hr light:12 hr

dark cycle and given ad libitum access to food and water. The Washington

University in St. Louis Animal Studies IACUC approved all experimental

methods.

Viral Injections and Surgical Procedures for Fiber Optic Placement

All surgeries were performed under isoflurane anesthesia (Piramal Healthcare,

Maharashtra, India). For slice physiology experiments, adult mice were

injected bilaterally with AAV5-EF1a-DIO-opto-MOR-eYFP. For behavior

experiments, adult male mice were injected unilaterally with 500 nl of AAV5-

EF1a-DIO-opto-MOR-eYFP virus (WUSTL Hope Center Viral Core, St. Louis,

MO) into the RMTg (AP: �3.9 mm, ML: �0.1 mm, DV: �4.5 mm) or the VP

(AP: 0.7 mm, ML: 1 mm, DV: �5.25 mm). Fiber optic ferrules were chronically

implanted above the VTA (AP: �3.4 mm, ML: �0.5 mm, DV: �4.0 mm) or the

VP (AP: 0.7 mm, ML: 1 mm, DV: �4.8 mm), respectively, and dental cement

(Lang Dental, Wheeling, IL) was applied to hold the ferrules in place (Sparta

et al., 2012). Mice were allowed to recover for at least 3 weeks following infu-

sion of virus and fiber placement prior to further behavioral testing.

Slice Physiology

Acute brain slices were prepared using a protective cutting and recovery

method (Ting et al., 2014). Anesthetized mice were transcardially perfused

with NMDG-substituted aCSF containing (in mM) 93 NMDG, 2.5 KCl, 1.25

NaH2PO4, 30 NaHCO3, 20 HEPES, 25 glucose, 5 ascorbic acid, 2 thiourea,

3 Na-pyruvate, 12 N-acetyl-L-cysteine, 10 MgSO4, 0.5 CaCl2 (pH = 7.3).

200 mm thick coronal sections of the PAG were cut using a Vibratome

VT1000s (Leica) and transferred to an oxygenated recovery chamber contain-

ing NMDGaCSF for 5–10min at 32�C –34�Cbefore being transferred to a hold-

ing chamber filled with modified aCSF containing (in mM) 92 NaCl, 2.5 KCl, 1.2

NaH2PO4, 30 NaHCO3, 20 HEPES, 25 glucose, 2 CaCl2, 2 MgCl2 (pH adjusted

to 7.3 with NaOH), and Osm 300-315 and incubated in the dark at room

temperature for >1 hr before recording. 225 mm thick sagittal sections were

prepared for experiments in RMTg. Additional details can be found in the

Supplemental Experimental Procedures.

Behavior

Real-Time Place Preference and Aversion, rtPP/A

Mice were placed into a custom-made black acrylic two-chambered box

(52.5 3 25.5 3 25.5 cm) and allowed to explore each of two chambers for

60 min. Using Noldus Ethovision hardware controller connected to a master

9 functional generator, light stimulation (473 nm, 10 mW) was delivered

through fiber optic implants during the duration of their time spent in the

‘‘conditioned’’ side of the chamber, and mice received no stimulation on the

‘‘unconditioned’’ side. The experimental animals were counterbalanced for

both group and conditioning side. Preference and aversion scores in each

experiment were determined by comparing the amount of time spent in the

conditioned versus unconditioned sides during the real-time testing phase.

Preference (RMTg) and aversion (VP) were calculated by comparing time spent

in the light stimulation box with time spent in the no stimulation box during the

final 30 min of the test session. All behavioral data (locomotion, velocity,

entries) were analyzed using Noldus Ethovision (v9.5) with a ZR900 Canon

camera.

Immunohistochemistry

As previously described (Kim et al., 2013), after the conclusion of behavioral

testing, mice were anesthetized with sodium pentobarbital and transcardially

perfused with ice cold PBS, followed by 4% phosphate-buffered paraformal-

dehyde. Brains were removed, post-fixed overnight in paraformaldehyde, and

saturated in 30% phosphate-buffered sucrose. 50 mm sections were cut,

washed in 0.3% Triton X-100/5% normal goat serum in 0.1 M PBS, stained

with fluorescent Nissl stain (1:400 Neurotrace, Invitrogen) for 1 hr, and

mounted onto glass slides with Vectashield (Vector Laboratories). VTA sec-

tions were stained with rabbit anti-tyrosine hydroxylase (1:1,000, Millipore)

overnight at 4�C and AlexaFluor 633 goat anti-rabbit for 2 hr at room temper-

ature (1:1,000, Molecular Probes) prior to the Nissl step. Opto-MOR expres-

sion was verified using fluorescence (Olympus) and confocal microscopy

(Leica Microsystems). Images were produced with Leica Application Suite

Advanced Fluorescence software. Animals that did not show targeted expres-

sion were excluded.

Neuron 86, 923–935, May 20, 2015 ª2015 Elsevier Inc. 933

SUPPLEMENTAL INFORMATION

Supplemental Information includes Supplemental Experimental Procedures

and five figures and can be found with this article online at http://dx.doi.org/

10.1016/j.neuron.2015.03.066.

AUTHOR CONTRIBUTIONS

E.R.S., B.A.C., and M.J.S. designed and performed experiments, collected

and analyzed data, and wrote the manuscript. M.A.B., R.A., S.C.F., and

J.G.M. designed and performed experiments and collected and analyzed

data. W.J.P. provided technical support and facilitated design, construction,

and cloning of the receptor chimera and virus. R.W.G. helped design and over-

see experiments. M.R.B. helped design, analyze, and oversee experiments

and wrote the manuscript.

ACKNOWLEDGMENTS

This work is supported by EUREKA NIDA R01DA037152 (to M.R.B.), NIMH

F31MH101956 (J.G.M.), Transformative Research Award NS081707 (to

R.W.G. and M.R.B.), NIDA K99DA038725 (to R.A.), the W.M. Keck Fellowship

in Molecular Medicine (M.J.S. and B.A.C.), and TR32 GM108539 (B.A.C.).

M.A.B. was supported by the Howard Hughes Medical Institute. We would

like to thank Vijay Samineni and Clint Morgan in the Gereau lab for their assis-

tance. We would also like to thank the Hope Center Viral Core at Washington

University-St. Louis. We thank Skylar Spangler, Lamley Lawson, and other

members of the Bruchas laboratory for their help revising the manuscript

and technical support. We also thank the laboratory of Dr. Evan Kharasch

for their technical support.

Received: November 6, 2014

Revised: March 6, 2015

Accepted: March 29, 2015

Published: April 30, 2015

REFERENCES

Airan, R.D., Thompson, K.R., Fenno, L.E., Bernstein, H., and Deisseroth, K.

(2009). Temporally precise in vivo control of intracellular signalling. Nature

458, 1025–1029.

Al-Hasani, R., and Bruchas, M.R. (2011). Molecular mechanisms of opioid re-

ceptor-dependent signaling and behavior. Anesthesiology 115, 1363–1381.

Atasoy, D., Aponte, Y., Su, H.H., and Sternson, S.M. (2008). A FLEX switch tar-

gets Channelrhodopsin-2 to multiple cell types for imaging and long-range cir-

cuit mapping. J. Neurosci. 28, 7025–7030.

Banghart, M.R., and Sabatini, B.L. (2012). Photoactivatable neuropeptides for

spatiotemporally precise delivery of opioids in neural tissue. Neuron 73,

249–259.

Barish, P.A., Xu, Y., Li, J., Sun, J., Jarajapu, Y.P.R., and Ogle, W.O. (2013).

Design and functional evaluation of an optically active m-opioid receptor.

Eur. J. Pharmacol. 705, 42–48.

Belcheva, M.M., Clark, A.L., Haas, P.D., Serna, J.S., Hahn, J.W., Kiss, A., and

Coscia, C.J. (2005). Mu and kappa opioid receptors activate ERK/MAPK via

different protein kinase C isoforms and secondary messengers in astrocytes.

J. Biol. Chem. 280, 27662–27669.

Berndt, A., Lee, S.Y., Ramakrishnan, C., and Deisseroth, K. (2014). Structure-

guided transformation of channelrhodopsin into a light-activated chloride

channel. Science 344, 420–424.

Blanchet, C., and Luscher, C. (2002). Desensitization of mu-opioid receptor-

evoked potassium currents: initiation at the receptor, expression at the

effector. Proc. Natl. Acad. Sci. USA 99, 4674–4679.

Bruchas, M.R., Schindler, A.G., Shankar, H., Messinger, D.I., Miyatake, M.,

Land, B.B., Lemos, J.C., Hagan, C.E., Neumaier, J.F., Quintana, A., et al.

(2011). Selective p38aMAPKdeletion in serotonergic neurons produces stress

resilience in models of depression and addiction. Neuron 71, 498–511.

934 Neuron 86, 923–935, May 20, 2015 ª2015 Elsevier Inc.

Celver, J., Xu,M., Jin,W., Lowe, J., and Chavkin, C. (2004). Distinct domains of

the m-opioid receptor control uncoupling and internalization. Mol. Pharmacol.

65, 528–537.

Childers, S.R., and Snyder, S.H. (1978). Guanine nucleotides differentiate

agonist and antagonist interactions with opiate receptors. Life Sci. 23,

759–761.

da Silva, S., Hasegawa, H., Scott, A., Zhou, X., Wagner, A.K., Han, B.-X., and

Wang, F. (2011). Proper formation of whisker barrelettes requires periphery-

derived Smad4-dependent TGF-beta signaling. Proc. Natl. Acad. Sci. USA

108, 3395–3400.

Gunaydin, L.A., Grosenick, L., Finkelstein, J.C., Kauvar, I.V., Fenno, L.E.,

Adhikari, A., Lammel, S., Mirzabekov, J.J., Airan, R.D., Zalocusky, K.A.,

et al. (2014). Natural neural projection dynamics underlying social behavior.

Cell 157, 1535–1551.

Gupta, A., Gomes, I., Wardman, J., andDevi, L.A. (2014). Opioid receptor func-

tion is regulated by post-endocytic peptide processing. J. Biol. Chem. 289,

19613–19626.

Hille, B. (1994). Modulation of ion-channel function by G-protein-coupled re-

ceptors. Trends Neurosci. 17, 531–536.

Hjelmstad, G.O., Xia, Y., Margolis, E.B., and Fields, H.L. (2013). Opioid modu-

lation of ventral pallidal afferents to ventral tegmental area neurons.

J. Neurosci. 33, 6454–6459.

Ingram, S.L., Macey, T.A., Fossum, E.N., and Morgan, M.M. (2008). Tolerance

to repeated morphine administration is associated with increased potency of

opioid agonists. Neuropsychopharmacology 33, 2494–2504.

Irannejad, R., Tomshine, J.C., Tomshine, J.R., Chevalier, M., Mahoney, J.P.,

Steyaert, J., Rasmussen, S.G.F., Sunahara, R.K., El-Samad, H., Huang, B.,

and von Zastrow, M. (2013). Conformational biosensors reveal GPCR signal-

ling from endosomes. Nature 495, 534–538.

Jalabert, M., Bourdy, R., Courtin, J., Veinante, P., Manzoni, O.J., Barrot, M.,

and Georges, F. (2011). Neuronal circuits underlying acute morphine action

on dopamine neurons. Proc. Natl. Acad. Sci. USA 108, 16446–16450.

Jennings, J.H., Sparta, D.R., Stamatakis, A.M., Ung, R.L., Pleil, K.E., Kash,

T.L., and Stuber, G.D. (2013). Distinct extended amygdala circuits for diver-

gent motivational states. Nature 496, 224–228.

Jhou, T.C., Xu, S.-P., Lee, M.R., Gallen, C.L., and Ikemoto, S. (2012). Mapping

of reinforcing and analgesic effects of the mu opioid agonist endomorphin-1 in

the ventral midbrain of the rat. Psychopharmacology (Berl.) 224, 303–312.

Kim, J.-M., Hwa, J., Garriga, P., Reeves, P.J., RajBhandary, U.L., and

Khorana, H.G. (2005). Light-driven activation of b 2-adrenergic receptor

signaling by a chimeric rhodopsin containing the b 2-adrenergic receptor cyto-

plasmic loops. Biochemistry 44, 2284–2292.

Kim, T.I., McCall, J.G., Jung, Y.H., Huang, X., Siuda, E.R., Li, Y., Song, J.,

Song, Y.M., Pao, H.A., Kim, R.-H., et al. (2013). Injectable, cellular-scale opto-

electronics with applications for wireless optogenetics. Science 340, 211–216.

Li, X., Gutierrez, D.V., Hanson,M.G., Han, J.,Mark, M.D., Chiel, H., Hegemann,

P., Landmesser, L.T., and Herlitze, S. (2005). Fast noninvasive activation and

inhibition of neural and network activity by vertebrate rhodopsin and green

algae channelrhodopsin. Proc. Natl. Acad. Sci. USA 102, 17816–17821.

Luscher, C., Jan, L.Y., Stoffel, M., Malenka, R.C., and Nicoll, R.A. (1997). G

protein-coupled inwardly rectifying K+ channels (GIRKs) mediate postsynaptic

but not presynaptic transmitter actions in hippocampal neurons. Neuron 19,

687–695.

Masseck, O.A., Spoida, K., Dalkara, D., Maejima, T., Rubelowski, J.M.,

Wallhorn, L., Deneris, E.S., and Herlitze, S. (2014). Vertebrate cone opsins

enable sustained and highly sensitive rapid control of Gi/o signaling in anxiety

circuitry. Neuron 81, 1263–1273.

Matsui, A., and Williams, J.T. (2011). Opioid-sensitive GABA inputs from ros-

tromedial tegmental nucleus synapse onto midbrain dopamine neurons.

J. Neurosci. 31, 17729–17735.

Matsui, A., Jarvie, B.C., Robinson, B.G., Hentges, S.T., and Williams, J.T.

(2014). Separate GABA afferents to dopamine neurons mediate acute action

of opioids, development of tolerance, and expression of withdrawal. Neuron

82, 1346–1356.

Raimondo, J.V., Kay, L., Ellender, T.J., and Akerman, C.J. (2012). Optogenetic

silencing strategies differ in their effects on inhibitory synaptic transmission.

Nat. Neurosci. 15, 1102–1104.

Roy, A., Kucukural, A., and Zhang, Y. (2010). I-TASSER: a unified platform for

automated protein structure and function prediction. Nat. Protoc. 5, 725–738.

Smith, K.S., and Berridge, K.C. (2005). The ventral pallidum and hedonic

reward: neurochemical maps of sucrose ‘‘liking’’ and food intake.

J. Neurosci. 25, 8637–8649.

Smith, K.S., Tindell, A.J., Aldridge, J.W., and Berridge, K.C. (2009). Ventral pal-

lidum roles in reward and motivation. Behav. Brain Res. 196, 155–167.

Sparta, D.R., Stamatakis, A.M., Phillips, J.L., Hovelsø, N., van Zessen, R., and

Stuber, G.D. (2012). Construction of implantable optical fibers for long-term

optogenetic manipulation of neural circuits. Nat. Protoc. 7, 12–23.

Tan, K.R., Yvon, C., Turiault, M., Mirzabekov, J.J., Doehner, J., Labouebe, G.,

Deisseroth, K., Tye, K.M., and Luscher, C. (2012). GABA neurons of the VTA

drive conditioned place aversion. Neuron 73, 1173–1183.

Ting, J.T., Daigle, T.L., Chen, Q., and Feng, G. (2014). Acute brain slice

methods for adult and aging animals: application of targeted patch clamp anal-

ysis and optogenetics. Methods Mol. Biol. 1183, 221–242.

Tsvetanova, N.G., and von Zastrow, M. (2014). Spatial encoding of cyclic AMP

signaling specificity by GPCR endocytosis. Nat. Chem. Biol. 10, 1061–1065.

van Zessen, R., Phillips, J.L., Budygin, E.A., and Stuber, G.D. (2012). Activation

of VTA GABA neurons disrupts reward consumption. Neuron 73, 1184–1194.

Vaughan, C.W., Bagley, E.E., Drew, G.M., Schuller, A., Pintar, J.E., Hack, S.P.,

and Christie, M.J. (2003). Cellular actions of opioids on periaqueductal grey

neurons from C57B16/J mice and mutant mice lacking MOR-1. Br. J.

Pharmacol. 139, 362–367.

Vong, L., Ye, C., Yang, Z., Choi, B., Chua, S., Jr., and Lowell, B.B. (2011).

Leptin action on GABAergic neurons prevents obesity and reduces inhibitory

tone to POMC neurons. Neuron 71, 142–154.

Wietek, J., Wiegert, J.S., Adeishvili, N., Schneider, F., Watanabe, H., Tsunoda,

S.P., Vogt, A., Elstner, M., Oertner, T.G., and Hegemann, P. (2014). Conversion

of channelrhodopsin into a light-gated chloride channel. Science 344,

409–412.

Zhang, F., Gradinaru, V., Adamantidis, A.R., Durand, R., Airan, R.D., de Lecea,

L., andDeisseroth, K. (2010). Optogenetic interrogation of neural circuits: tech-

nology for probing mammalian brain structures. Nat. Protoc. 5, 439–456.

Zhou, X., Wang, L., Hasegawa, H., Amin, P., Han, B.-X., Kaneko, S., He, Y.,

and Wang, F. (2010). Deletion of PIK3C3/Vps34 in sensory neurons causes

rapid neurodegeneration by disrupting the endosomal but not the autophagic

pathway. Proc. Natl. Acad. Sci. USA 107, 9424–9429.

Neuron 86, 923–935, May 20, 2015 ª2015 Elsevier Inc. 935